A multiparametric risk table for loss of clinical remission status in patients with rheumatoid arthritis: a STARTER study post-hoc analysis

Abstract

Abstract Objective This post-hoc analysis was carried out on data acquired in the longitudinal Sonographic Tenosynovitis Assessment in RheumaToid arthritis patiEnts in Remission (STARTER) study. Our primary aim was to determine the predictive clinical and musculoskeletal ultrasonographic (MSUS) features associated with disease flare in RA patients in clinical remission, while our secondary aim was to evaluate the probability of disease flare based on clinical and MSUS features. Methods We analysed data for a total of 389 RA patients in DAS28-defined remission. All patients underwent a MSUS examination according to the OMERACT guidelines. Logistic regression and results, presented as odds ratio and 95% CI, were used for the evaluation of the association between selected variables and disease flare. Significant clinical and MSUS features were incorporated into a risk table for predicting disease flare within at least 12 months of follow-up in patients with RA remission. Results Within 12 months, 137 (35%) RA patients experienced a disease flare. RA patients who experienced a flare disease differed from those with persistent remission in terms of ACPA positivity (75.9% vs 62.3%, respectively; P = 0.007), percentage of sustained clinical remission at baseline (44.1% vs 68.5%, respectively; P = 0.001) and synovium power Doppler signal presence (58.4% vs 33.3%, respectively; P < 0.001). Based on these results, these three features were considered in a predictive model of disease flare with an adjusted odds ratio of 3.064 (95% CI 1.728–5.432). Finally, a risk table was constructed including the three significant predictive factors of disease flare occurring within 12 months from the enrolment. Conclusion An adaptive flare-prediction model tool, based on data available in outpatient settings, was developed as a multiparametric risk table. If confirmed by external validation, this tool might support the defining of therapeutic strategies in RA patients in DAS28-defined remission status.